US7839777B2 - Method, system, and apparatus for accelerating resolution of network congestion - Google Patents

Method, system, and apparatus for accelerating resolution of network congestion Download PDF

Info

Publication number
US7839777B2
US7839777B2 US11/862,370 US86237007A US7839777B2 US 7839777 B2 US7839777 B2 US 7839777B2 US 86237007 A US86237007 A US 86237007A US 7839777 B2 US7839777 B2 US 7839777B2
Authority
US
United States
Prior art keywords
notification signal
data packet
congestion notification
backward
node
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active, expires
Application number
US11/862,370
Other versions
US20090086637A1 (en
Inventor
Casimer DeCusatis
Thomas A. Gregg
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
International Business Machines Corp
Original Assignee
International Business Machines Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by International Business Machines Corp filed Critical International Business Machines Corp
Priority to US11/862,370 priority Critical patent/US7839777B2/en
Assigned to INTERNATIONAL BUSINESS MACHINES CORPORATION reassignment INTERNATIONAL BUSINESS MACHINES CORPORATION ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: DECUSATIS, CASIMER, GREGG, THOMAS A.
Publication of US20090086637A1 publication Critical patent/US20090086637A1/en
Application granted granted Critical
Publication of US7839777B2 publication Critical patent/US7839777B2/en
Application status is Active legal-status Critical
Adjusted expiration legal-status Critical

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic regulation in packet switching networks
    • H04L47/10Flow control or congestion control
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic regulation in packet switching networks
    • H04L47/10Flow control or congestion control
    • H04L47/11Congestion identification
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic regulation in packet switching networks
    • H04L47/10Flow control or congestion control
    • H04L47/26Explicit feedback to the source, e.g. choke packet
    • H04L47/263Source rate modification after feedback

Abstract

The response time for resolving network traffic congestion is accelerated in a Data Center Ethernet (DCE) network. A data packet is received at a node in the network. Congestion of the data packet at the node is detected, and a backward congestion notification signal for the data packet is generated. A packet injection rate is adapted based on at least one of the backward congestion notification signal generated by the node and another backward congestion notification signal.

Description

BACKGROUND

The present invention relates generally to network congestion, and, more particularly, to resolving network congestion.

Data Center Ethernet (DCE) is an emerging industry standard which proposes modification is to existing networks, in an effort to position Ethernet as the preferred convergence fabric or all types of data center traffic. A recent study has found that Ethernet is the convergence fabric, with I/O consolidation in a Data Center as shown in FIG. 1. This consolidation is expected to simplify platform architecture and reduce overall platform costs. More details of proposals for consolidation are described in “Proposal for Traffic Differentiation in Ethernet Networks,” which may be found at http://www.ieee802.org/1/files/public/docs2005/new-wadekar-virtual%20-links-0305.pdf.

Major changes have been proposed for DCE (also referred to as enhanced Ethernet and low latency Ethernet), including the addition of credit based flow control at the link layer, congestion detection and data rate throttling, and the addition of virtual lanes with quality of service differentiation. It is important to note that these functions do not affect Transmission Control Protocol/Internet Protocol (TCP/IP), which exists above the DCE level. It should also be noted that DCE is intended to operate without necessitating the overhead of TCP/IP. This offers a much simpler, low cost approach that does not require offload processing or accelerators.

An existing method for backward congestion control management may be understood with reference to FIG. 2. In the system shown in FIG. 2, a source node 220 transmits data packets to a destination node via a switch 210. The packets are transmitted via a virtual lane. When congestion occurs on the virtual lane for a given port, the switch 210 detects the congestion using a threshold comparison. This may involve measuring the volume of packets accumulated in the buffer of the switch and measuring an arrival rate of packet sequence numbers. Further details of packet sequence numbers are provided in commonly assigned U.S. patent application Ser. Nos. 11/847,965 and 11/426,421, herein incorporated by reference.

When congestion is detected by the switch 210, the switch sets an explicit congestion notification (ECN) bit in the header of the appropriate data packet being transmitted downstream towards the destination node 230. Upon receipt of a data packet including an ECN bit on, the destination node 230 sets a backward explicit congestion notification (BECN) bit on in the data packet header and transmits the data packet upstream towards the source node 220. When the source node 220 receives the data packet with the BECN bit on, the injection rate of data packets sent downstream is reduced. The source node 220 maintains a table 225 of inter-packet delays used for injection rate control. Each time a data packet is received at the source node 220 with a BECN on, the source node decrements the injection rate to the next lowest rate in the table, i.e., the source node uses the next longer delay in the table for injection rate control. The table index is decremented based on a timer. Whenever the timer expires without receipt of any additional data packets with BECN bits on, the source data rate is allowed to increase by one increment. Eventually, if no more data packets are received with BECN bit on, the injection rate fully recovers.

One problem with this approach is that the source data rate is only decreased after data packets including the ECN and the BECN have made almost one full round trip through the network. In order to accommodate the resulting travel time delays, the nodes must be designed with sufficiently large buffers. The buffer size, and associated node cost, can become quite large when it is desirable for the network to operate at extended distances (e.g., 10 s of kilometers, which is typical for disaster recovery applications). This type of congestion management places a fundamental limit on the maximum distance that can be achieved in a DCE network. Since DCE networks are not necessarily point-to-point, but may involve numerous hops and switch cascades, long cable distances may accumulate even if the source and destination are not geographically distributed far from each other.

Furthermore, the conventional congestion control approach is essentially a long delay feedback loop. Because it takes so long for the ECN/BECN notifications to reach their destinations, the feedback loop is not able to stay current with the state of the network. In fact, the feedback loop may lag behind other forms of credit based flow controls at the link layer.

For at least these reasons, it is desirable to reduce the response time of a backward congestion control scheme and thereby allow for longer distance links without excessively large buffer credits.

SUMMARY

According to exemplary embodiments, a method, a system, and an apparatus are provided for managing backward network congestion to accelerate resolution of congestion in a Data Center Ethernet (DCE) network. A data packet is received at a node in the network. Congestion of the data packet at the node is detected, and a backward congestion notification signal for the data packet is generated. A packet injection rate is adapted based on at least one of the backward congestion notification signal generated by the node and another backward congestion notification signal.

According to an exemplary embodiment, the data packet is received at the node from a source node. The other backward congestion notification signal is generated by a destination node and transmitted for the data packet to the source node in response to receipt of a congestion notification signal transmitted with the data packet from the other node. The source node adapts the packet injection rate based on the backward congestion signal received first. If one or more backward notification signals are received for the same data packet, duplication of the backward notification signal is detected by comparison of packet sequence numbers (PSNs) included in data headers with the backward notification congestion signals. Duplicate backward congestion notification signals received subsequently for the data packet are ignored.

BRIEF DESCRIPTION OF THE DRAWINGS

Referring to the exemplary drawings, wherein like elements are numbered alike in the several Figures:

FIG. 1 illustrates proposed consolidation of traffic in a Data Center Ethernet (DCE) network;

FIG. 2 illustrates how a conventional DCE network resolves congestion;

FIG. 3 illustrates how congestion is resolved in a DCE network according to an exemplary embodiment;

FIG. 4 illustrates the flow of congestion control in a DCE network according to an exemplary embodiment.

DETAILED DESCRIPTION

According to an exemplary embodiment, when a switch or other node in a DCE network detects port congestion, a congestion bit may be toggled on in both the forward and backward propagating data streams. The switch may be modified so that a DCE packet can be inserted in the backward propagating direction, with the same address as the source of the virtual lane causing the congestion. The backward propagating packet may be a copy of the packet that is to be echoed from the destination node back to the source node and may be assigned the same PSN as the forward propagating packet. Due to variable delays encountered by packets traversing the network, it is not clear whether the forward or backward propagating packet will arrive back at the source first. However, the use of PSNs allows packet duplication to be detected. Thus, the packet having a particular PSN that arrives first at the source with its BECN bit set on will adapt the packet injection rate, i.e., initiate throttling down of the packet transmission rate. A packet that arrives subsequently with a BECN bit set on but with the same PSN of the packet that arrived first will be ignored. Thus, multiple packets having the same PSN with the BECN bit set on may be sent back to the source, such that the source is made aware of congestion as fast as possible. The source will just ignore packets having the same PSN as a packet first received, so the source does not become overwhelmed with requests to throttle down the packet transmission rate based on the same packet.

Because of the ability to identify packets in flight, in either direction across other networks, multiple congestion notifications can exist in both directions through the fabric and can resolve themselves at the source node or destination node. According to an exemplary embodiment, both the congestion point (e.g., switch) and the receiving end point (destination node) send a packet with the BECN bit set on.

FIG. 3 illustrates how congestion is resolved in a DCE network according to an exemplary embodiment. As shown in FIG. 3, a packet with a PSN is sent to a switch 310 from a source node 320. As indicated in FIG. 3, congestion is detected in the switch 310. Congestion may be detected any manner, including that described in commonly assigned, copending U.S. patent application Ser. No. 11/861,564, filed on Sep. 26, 2007, entitled “Method, System, and Computer Program Product for Adaptive Congestion Control on Virtual Lanes for Data Center Ethernet Architecture,” and hereby incorporated by reference.

An ECN bit is set on in the header of the data packet sent to the destination node 320, and the destination node responds to the switch 320 with a BECN bit and a PSN identifying the packet. The switch, in turn, transmits the BECN bit and PSN from the destination to the source.

As the switch transmits an ECN bit in the header of the data packet to the destination 330, a packet generator in the switch 310 generates a BECN and a PSN identifying the packet. As those skilled in the art will appreciate, the packet generator may be implemented with customized hardware and/or software, built as desired. The BECN and PSN generated by the packet generator may be formed as a packet header that is substantially a copy of the data packet header sent to the destination node, with a BECN bit set on in the header. The BECN and PSN generated by the switch are sent to the source node 320.

If the source node 320 receives both the BECN generated by the switch and the BECN generated by the destination node 330, the source node 320 reacts to the BECN received first to adapt the packet injection rate, i.e., throttle down the packet transmission rate. If the congested switch is located close to the source node 320, the BECN generated by the switch for a particular packet may be received first. In this case, when the source node 320 receives the BECN for the same packet from the destination node 330, the source node performs a PSN comparison and detects that the BECN bit from the destination node is a duplicate. Thus, the source node ignores the BECN from the destination node. Similarly, if the BECN for a particular packet is received first from the destination node, the BECN for the same packet received from the switch is ignored.

This backward congestion management technique can be significantly faster at throttling down the source data transmission rate compared with prior approaches. The delay between congestion detection and source throttling may be substantially cut compared with prior approaches.

In addition, if one of the BECNs is lost, the source node 320 reacts to the other BECN as soon as it is received. The redundancy in BECN generation provides for fault-tolerant congestion control.

FIG. 4 illustrates a flow of congestion control in a DCE network according to an exemplary embodiment. In this embodiment, data packets are transmitted from source node 320 to a destination node 330 via switches 310 a and 310 b and a DCE fabric of links 340. Although only two switches are shown for simplicity of illustration, it should be appreciated that there may be any number of switches or nodes that may detect congestion. In the example shown in FIG. 4, the switch 310 a detects congestion of a data packet, sets an ECN bit on in a data packet header and transmit the data packet with the ECN bit set on to the destination node via the DCE fabric 340 and the switch 310 b as explained above. It should be appreciated, however, that congestion may be detected at other places, as well, such as nodes within the fabric 340, at the switch 310 b, or at the destination node 330. As the ECNs are received at the fabric 340, the switch 310 b, and the destination node 330, each of these components generate a BECN. According to an exemplary embodiment, the data packet header may include, in addition to the BECN, a field that identifies the source of the BECNs, and multiple BECNs may be received by the source node 320. For example, as illustrated in FIG. 4, BECN(1) is a backward congestion bit set on by the switch 310 a, BECN(2) is a backward congestion bit set on by a switch or node in the DCE fabric 340, BECN(3) is a backward congestion bit set on by the switch 310 b, and BECN(4) is a backward congestion bit set on the by destination node 330. By identifying the source of the BECN, mapping of network behavior is made possible. That is, when the source node 320 receives the BECNs, it can determine the source that transmitted the BECN based on the field identifying the BECN and thus determine at what point(s) congestion is occurring.

As in the system shown in FIG. 3, when the source node 320 receives a BECN for a data packet, the source throttles down packet transmission rate. If the source node 320 determines by PSN comparison that multiple BECNs are received subsequently for the same data packet, the source just ignores the subsequently received BECNs.

Although FIGS. 3 and 4 illustrate congestion detection and BECN generation in switches, it may be desirable to incorporate this architecture into channel extenders of wavelength division multiplexing (WDM) equipment specially designed for use with DCE extended distance links. Since the proposed approach is also fully compatible with legacy backward congestion control, and the two can co-exist in the same network, this backward management congestion control capability may not be needed in every node in a DCE network to obtain benefits.

While the invention has been described with reference to exemplary embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the scope of the invention. In addition, many modifications may be made to adapt a particular situation or material to the teachings of the invention without departing from the essential scope thereof. Therefore, it is intended that the invention not be limited to the particular embodiment disclosed as the best mode contemplated for carrying out this invention, but that the invention will include all embodiments falling within the scope of the appended claims.

Claims (15)

1. A method for managing backward congestion in a Data Center Ethernet (DCE) network, comprising:
receiving a data packet at a node in the network;
detecting congestion of the data packet at the node in the network;
generating at least one backward congestion notification signal for the data packet at the node in the network, wherein a packet injection rate is adapted based on at least one of the backward congestion notification signal generated by the node in the network and another backward congestion notification signal;
wherein the data packet is received at the node in the network from a source node, the another backward congestion notification signal is generated by a destination node and transmitted for the data packet to the source node in response to receipt of a congestion notification signal transmitted with the data packet from the node in the network, and the source node adapts the packet injection rate based on the backward congestion signal received first;
wherein the congestion notification signal is a data bit included in a data packet header of the data packet transmitted to the destination node, and each backward congestion notification signal is a data bit included in a data packet header transmitted to the source node;
wherein each data packet header including a backward congestion notification signal also includes a field indicating a source of the backward congestion notification signal.
2. A method for managing backward congestion in a Data Center Ethernet (DCE) network, comprising:
receiving a data packet at a node in the network;
detecting congestion of the data packet at the node in the network;
generating at least one backward congestion notification signal for the data packet at the node in the network, wherein a packet injection rate is adapted based on at least one of the backward congestion notification signal generated by the node in the network and another backward congestion notification signal;
wherein the data packet is received at the node in the network from a source node, the another backward congestion notification signal is generated by a destination node and transmitted for the data packet to the source node in response to receipt of a congestion notification signal transmitted with the data packet from the node in the network, and the source node adapts the packet injection rate based on the backward congestion signal received first;
wherein each data packet header includes a packet sequence number (PSN) for that data packet.
3. The method of claim 2, wherein the congestion notification signal is a data bit included in a data packet header of the data packet transmitted to the destination node, and each backward congestion notification signal is a data bit included in a data packet header transmitted to the source node.
4. The method of claim 2, wherein if one or more backward congestion notification signals are received for the same data packet, duplication of the backward congestion notification signal is detected by comparison of the PSNs included in the data headers with the backward congestion notification signals.
5. The method of claim 4, further comprising ignoring a duplicate backward congestion notification signal received subsequently for the data packet.
6. A system for managing backward network congestion in a Data Center Ethernet (DCE) network, comprising:
a source node for transmitting data packets;
a destination node for receiving the data packets; and
at least one node interspersed between the source node and the destination node for receiving a data packet from the source node, detecting congestion of the data packet in the network, generating a congestion notification signal indicating congestion of the data packet, and transmitting the data packet with the congestion notification signal to a destination node in the network;
wherein the node generates a backward congestion notification signal and transmits the backward congestion notification signal to the source node;
the destination node receives the congestion notification signal, generates another backward congestion notification signal and transmits the another backward congestion notification signal to the source node; and
the source node receives at least one of the backward congestion notification signals and adapts a packet injection rate based on the backward congestion notification signal received first;
wherein the congestion notification signal is a data bit included in a data packet header of the data packet transmitted to the destination node;
wherein each backward congestion notification signal is a data bit included in a data packet header transmitted to the source node;
wherein the data packet header including the backward congestion notification signal also includes a field indicating a source of the backward congestion notification signal.
7. A system for managing backward network congestion in a Data Center Ethernet (DCE) network, comprising:
a source node for transmitting data packets;
a destination node for receiving the data packets; and
at least one node interspersed between the source node and the destination node for receiving data packet from the source node, detecting congestion of the data packet in the network, generating a congestion notification signal indicating congestion of the data packet, and transmitting the data packet with the congestion notification signal to a destination node in the network;
wherein the node generates a backward congestion notification signal and transmits the backward congestion notification signal to the source node;
the destination node receives the congestion notification signal, generates another backward congestion notification signal and transmits the another backward congestion notification signal to the source node; and
the source node receives at least one of the backward congestion notification signals and adapts a packet injection rate based on the backward congestion notification signal received first;
wherein each data packet header includes a packet sequence number (PSN) for that data packet.
8. The system of claim 7, wherein the congestion notification signal is a data bit included in a data packet header of the data packet transmitted to the destination node.
9. The system of claim 7, wherein each backward congestion notification signal is a data bit included in a data packet header transmitted to the source node.
10. The system of claim 7, wherein if the source node subsequently receives one or more backward congestion notification signals for the same data packet, the source node detects duplication of the backward congestion notification signal by comparison of the PSNs included in the data headers with the backward congestion notification signals.
11. The system claim 10, wherein the source node ignores a duplicate backward congestion notification signal received subsequently for the data packet at the source node.
12. An apparatus for managing backward network congestion in a Data Center Ethernet (DCE) network, comprising:
an input for receiving a data packet;
a detector for detecting congestion of the data packet;
a packet generator for generating a backward congestion notification signal for the data packet, wherein a packet injection rate is adapted based on at least one of the backward congestion notification signal generated by the packet generator and another backward congestion notification signal;
an output for transmitting a congestion notification signal with the data packet to a destination node; and
another output for transmitting the backward congestion notification signal for the data packet to a source node, wherein the another backward congestion notification signal is generated by the destination node and transmitted for the data packet to the source node in response to receipt of the congestion notification signal, and the source node adapts the packet injection rate based on the backward congestion notification signal received first;
wherein the congestion notification signal is a data bit included in a header of the data packet transmitted to the destination node, and each backward congestion notification signal is a data bit included in a header of a data packet transmitted to the source node;
wherein each data packet header including a backward congestion notification signal also includes a field indicating a source of the backward congestion notification signal.
13. An apparatus for managing backward network congestion in a Data Center Ethernet (DCE) network, comprising:
an input for receiving a data packet;
a detector for detecting congestion of the data packet;
a packet generator for generating a backward congestion notification signal for the data packet, wherein a packet injection rate is adapted based on at least one of the backward congestion notification signal generated by the packet generator and another backward congestion notification signal;
an output for transmitting a congestion notification signal with the data packet to a destination node; and
another output for transmitting the backward congestion notification signal for the data packet to a source node, wherein the another backward congestion notification signal is generated by the destination node and transmitted for the data packet to the source node in response to receipt of the congestion notification signal, and the source node adapts the packet injection rate based on the backward congestion notification signal received first;
wherein each data packet header includes a packet sequence number (PSN) for that data packet.
14. The apparatus of claim 13, wherein the congestion notification signal is a data bit included in a header of the data packet transmitted to the destination node, and each backward congestion notification signal is a data bit included in a header of a data packet transmitted to the source node.
15. The apparatus of claim 13, wherein if one or more backward notification signals are received for the same data packet, duplication of the backward notification signal is detected by comparison of the PSNs included in the data headers with the backward congestion notification signals, and a duplicate backward congestion notification signal received subsequently for the data packet is ignored.
US11/862,370 2007-09-27 2007-09-27 Method, system, and apparatus for accelerating resolution of network congestion Active 2029-01-21 US7839777B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US11/862,370 US7839777B2 (en) 2007-09-27 2007-09-27 Method, system, and apparatus for accelerating resolution of network congestion

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US11/862,370 US7839777B2 (en) 2007-09-27 2007-09-27 Method, system, and apparatus for accelerating resolution of network congestion

Publications (2)

Publication Number Publication Date
US20090086637A1 US20090086637A1 (en) 2009-04-02
US7839777B2 true US7839777B2 (en) 2010-11-23

Family

ID=40508162

Family Applications (1)

Application Number Title Priority Date Filing Date
US11/862,370 Active 2029-01-21 US7839777B2 (en) 2007-09-27 2007-09-27 Method, system, and apparatus for accelerating resolution of network congestion

Country Status (1)

Country Link
US (1) US7839777B2 (en)

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050060423A1 (en) * 2003-09-15 2005-03-17 Sachin Garg Congestion management in telecommunications networks
US20110087943A1 (en) * 2009-10-12 2011-04-14 Empire Technology Development Llc Reliable communications in on-chip networks
US20110170408A1 (en) * 2010-01-11 2011-07-14 Research In Motion Limited Congestion level indication with explicit congestion notification in communication systems
US20110170410A1 (en) * 2010-01-11 2011-07-14 Research In Motion Limited Explicit congestion notification based rate adaptation using binary marking in communication systems
WO2013095664A1 (en) * 2011-12-23 2013-06-27 Intel Corporation Traffic control on an on-chip network
US8717874B2 (en) 2011-09-12 2014-05-06 International Business Machines Corporation Updating a switch software image in a distributed fabric protocol (DFP) switching network
US8750129B2 (en) 2011-10-06 2014-06-10 International Business Machines Corporation Credit-based network congestion management
US8767722B2 (en) 2011-05-14 2014-07-01 International Business Machines Corporation Data traffic handling in a distributed fabric protocol (DFP) switching network architecture
US8767529B2 (en) 2011-09-12 2014-07-01 International Business Machines Corporation High availability distributed fabric protocol (DFP) switching network architecture
US8798080B2 (en) 2011-05-14 2014-08-05 International Business Machines Corporation Distributed fabric protocol (DFP) switching network architecture
US8824485B2 (en) 2011-05-13 2014-09-02 International Business Machines Corporation Efficient software-based private VLAN solution for distributed virtual switches
US8856801B2 (en) 2011-05-14 2014-10-07 International Business Machines Corporation Techniques for executing normally interruptible threads in a non-preemptive manner
US8948004B2 (en) 2011-06-17 2015-02-03 International Business Machines Corporation Fault tolerant communication in a trill network
US9059922B2 (en) 2011-10-06 2015-06-16 International Business Machines Corporation Network traffic distribution
US9276953B2 (en) 2011-05-13 2016-03-01 International Business Machines Corporation Method and apparatus to detect and block unauthorized MAC address by virtual machine aware network switches
US9648560B2 (en) 2013-10-16 2017-05-09 At&T Intellectual Property I, L.P. Utilizing explicit congestion notification for network selection

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7719969B1 (en) * 2007-12-28 2010-05-18 Qlogic, Corporation System and method for assigning network device port address based on link rate
US8347031B2 (en) * 2009-04-14 2013-01-01 Hitachi, Ltd. Storage system and control method for managing the input and output of data in a network to avoid congestion
US9832125B2 (en) * 2015-05-18 2017-11-28 Dell Products L.P. Congestion notification system
US9807024B2 (en) * 2015-06-04 2017-10-31 Mellanox Technologies, Ltd. Management of data transmission limits for congestion control
US10009277B2 (en) 2015-08-04 2018-06-26 Mellanox Technologies Tlv Ltd. Backward congestion notification in layer-3 networks
US10237376B2 (en) 2015-09-29 2019-03-19 Mellanox Technologies, Ltd. Hardware-based congestion control for TCP traffic

Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5361259A (en) 1993-02-19 1994-11-01 American Telephone And Telegraph Company Wide area network (WAN)-arrangement
US5781532A (en) * 1993-12-23 1998-07-14 Newbridge Networks Corporation Data link interface for packet-switched network
US6418119B1 (en) 1997-09-22 2002-07-09 International Business Machines Corporation Data transmission apparatus and method thereof
US6426944B1 (en) 1998-12-30 2002-07-30 At&T Corp Method and apparatus for controlling data messages across a fast packet network
US6741555B1 (en) * 2000-06-14 2004-05-25 Nokia Internet Communictions Inc. Enhancement of explicit congestion notification (ECN) for wireless network applications
US20050144309A1 (en) 2003-12-16 2005-06-30 Intel Corporation, A Delaware Corporation Systems and methods for controlling congestion using a time-stamp
US7016971B1 (en) * 1999-05-24 2006-03-21 Hewlett-Packard Company Congestion management in a distributed computer system multiplying current variable injection rate with a constant to set new variable injection rate at source node
US20070268830A1 (en) * 2006-05-19 2007-11-22 Cisco Technology, Inc. Flow based flow control in an ethernet switch backplane
US20080025217A1 (en) * 2006-07-31 2008-01-31 Mircea Gusat System and Method For Enabling Management Of A Plurality Of Messages In A Communication Network
US20080259959A1 (en) * 2005-12-30 2008-10-23 Huawei Technologies Co., Ltd. Method for carrying frame relay over ethernet
US20090059916A1 (en) 2007-08-30 2009-03-05 International Business Machines Corporation Method, system, and apparatus for reliable data packet recovery in a link layer of a data center ethernet network
US20090080334A1 (en) 2007-09-26 2009-03-26 International Business Machines Corporation Method, system, and computer program product for adaptive congestion control on virtual lanes for data center ethernet architecture

Patent Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5361259A (en) 1993-02-19 1994-11-01 American Telephone And Telegraph Company Wide area network (WAN)-arrangement
US5781532A (en) * 1993-12-23 1998-07-14 Newbridge Networks Corporation Data link interface for packet-switched network
US6418119B1 (en) 1997-09-22 2002-07-09 International Business Machines Corporation Data transmission apparatus and method thereof
US6426944B1 (en) 1998-12-30 2002-07-30 At&T Corp Method and apparatus for controlling data messages across a fast packet network
US7016971B1 (en) * 1999-05-24 2006-03-21 Hewlett-Packard Company Congestion management in a distributed computer system multiplying current variable injection rate with a constant to set new variable injection rate at source node
US6741555B1 (en) * 2000-06-14 2004-05-25 Nokia Internet Communictions Inc. Enhancement of explicit congestion notification (ECN) for wireless network applications
US20050144309A1 (en) 2003-12-16 2005-06-30 Intel Corporation, A Delaware Corporation Systems and methods for controlling congestion using a time-stamp
US20080259959A1 (en) * 2005-12-30 2008-10-23 Huawei Technologies Co., Ltd. Method for carrying frame relay over ethernet
US20070268830A1 (en) * 2006-05-19 2007-11-22 Cisco Technology, Inc. Flow based flow control in an ethernet switch backplane
US20080025217A1 (en) * 2006-07-31 2008-01-31 Mircea Gusat System and Method For Enabling Management Of A Plurality Of Messages In A Communication Network
US20090059916A1 (en) 2007-08-30 2009-03-05 International Business Machines Corporation Method, system, and apparatus for reliable data packet recovery in a link layer of a data center ethernet network
US20090080334A1 (en) 2007-09-26 2009-03-26 International Business Machines Corporation Method, system, and computer program product for adaptive congestion control on virtual lanes for data center ethernet architecture

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
Proposal for Traffic Differentiation in Ethernet Networks, [online]; [retrieved on Sep. 20, 2007], retrieved from the Internet, http://www.ieee802.0rg/l/files/public/docs2005/new-wadekar-virtual%20-links-0305.pdf.

Cited By (28)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050060423A1 (en) * 2003-09-15 2005-03-17 Sachin Garg Congestion management in telecommunications networks
US20110087943A1 (en) * 2009-10-12 2011-04-14 Empire Technology Development Llc Reliable communications in on-chip networks
US8473818B2 (en) * 2009-10-12 2013-06-25 Empire Technology Development Llc Reliable communications in on-chip networks
US20110170408A1 (en) * 2010-01-11 2011-07-14 Research In Motion Limited Congestion level indication with explicit congestion notification in communication systems
US20110170410A1 (en) * 2010-01-11 2011-07-14 Research In Motion Limited Explicit congestion notification based rate adaptation using binary marking in communication systems
US8416690B2 (en) 2010-01-11 2013-04-09 Research In Motion Limited Explicit congestion notification based rate adaptation using binary marking in communication systems
US8693320B2 (en) * 2010-01-11 2014-04-08 Research In Motion Limited Congestion level indication with explicit congestion notification in communication systems
US9276953B2 (en) 2011-05-13 2016-03-01 International Business Machines Corporation Method and apparatus to detect and block unauthorized MAC address by virtual machine aware network switches
US8824485B2 (en) 2011-05-13 2014-09-02 International Business Machines Corporation Efficient software-based private VLAN solution for distributed virtual switches
US8856801B2 (en) 2011-05-14 2014-10-07 International Business Machines Corporation Techniques for executing normally interruptible threads in a non-preemptive manner
US8837499B2 (en) 2011-05-14 2014-09-16 International Business Machines Corporation Distributed fabric protocol (DFP) switching network architecture
US8767722B2 (en) 2011-05-14 2014-07-01 International Business Machines Corporation Data traffic handling in a distributed fabric protocol (DFP) switching network architecture
US8798080B2 (en) 2011-05-14 2014-08-05 International Business Machines Corporation Distributed fabric protocol (DFP) switching network architecture
US8948004B2 (en) 2011-06-17 2015-02-03 International Business Machines Corporation Fault tolerant communication in a trill network
US8948003B2 (en) 2011-06-17 2015-02-03 International Business Machines Corporation Fault tolerant communication in a TRILL network
US8767529B2 (en) 2011-09-12 2014-07-01 International Business Machines Corporation High availability distributed fabric protocol (DFP) switching network architecture
US8717874B2 (en) 2011-09-12 2014-05-06 International Business Machines Corporation Updating a switch software image in a distributed fabric protocol (DFP) switching network
US8797843B2 (en) 2011-09-12 2014-08-05 International Business Machines Corporation High availability distributed fabric protocol (DFP) switching network architecture
US8942094B2 (en) 2011-10-06 2015-01-27 International Business Machines Corporation Credit-based network congestion management
US8750129B2 (en) 2011-10-06 2014-06-10 International Business Machines Corporation Credit-based network congestion management
US9059922B2 (en) 2011-10-06 2015-06-16 International Business Machines Corporation Network traffic distribution
US9065745B2 (en) 2011-10-06 2015-06-23 International Business Machines Corporation Network traffic distribution
US20140010079A1 (en) * 2011-12-23 2014-01-09 Matteo Monchiero Traffic control on an on-chip network
WO2013095664A1 (en) * 2011-12-23 2013-06-27 Intel Corporation Traffic control on an on-chip network
US9608922B2 (en) * 2011-12-23 2017-03-28 Intel Corporation Traffic control on an on-chip network
TWI603205B (en) * 2011-12-23 2017-10-21 Intel Corp Traffic control on an on-chip network
US9648560B2 (en) 2013-10-16 2017-05-09 At&T Intellectual Property I, L.P. Utilizing explicit congestion notification for network selection
US9998979B2 (en) 2013-10-16 2018-06-12 At&T Intellectual Property I, L.P. Utilizing explicit congestion notification for network selection

Also Published As

Publication number Publication date
US20090086637A1 (en) 2009-04-02

Similar Documents

Publication Publication Date Title
Feng et al. A self-configuring RED gateway
Goyal et al. Improving the performance of TCP over the ATM-UBR service
US8411560B2 (en) TCP selection acknowledgements for communicating delivered and missing data packets
US7656799B2 (en) Flow control system architecture
EP1457008B1 (en) Methods and apparatus for network congestion control
US7698453B2 (en) Early generation of acknowledgements for flow control
US9008100B2 (en) Wavefront detection and disambiguation of acknowledgments
US6438101B1 (en) Method and apparatus for managing congestion within an internetwork using window adaptation
US8233392B2 (en) Transaction boundary detection for reduction in timeout penalties
US8873396B2 (en) Method and system for congestion marking
US6587471B1 (en) Methods, systems and computer program products for suppressing multiple destination traffic in a computer network
CN101803316B (en) Method, system, and computer program product for adaptive congestion control on virtual lanes for data center Ethernet architecture
US6535482B1 (en) Congestion notification from router
US20060056308A1 (en) Method of switching fabric for counteracting a saturation tree occurring in a network with nodes
Bohacek et al. A new TCP for persistent packet reordering
US20020172148A1 (en) Method and apparatus for protection switching in virtual private networks
EP1185127A2 (en) Reflection routing method in optical packet switching network and optical packet switch
JP4632874B2 (en) Communication terminal
US20020194361A1 (en) Data transmitting/receiving method, transmitting device, receiving device, transmiting/receiving system, and program
US9426195B2 (en) System and method for distribution of data packets utilizing an intelligent distribution network
US6192422B1 (en) Repeater with flow control device transmitting congestion indication data from output port buffer to associated network node upon port input buffer crossing threshold level
US6003089A (en) Method for constructing adaptive packet lengths in a congested network
US7839783B2 (en) Systems and methods of improving performance of transport protocols
US20100088437A1 (en) Infiniband adaptive congestion control adaptive marking rate
US8098648B2 (en) Load distributing method

Legal Events

Date Code Title Description
AS Assignment

Owner name: INTERNATIONAL BUSINESS MACHINES CORPORATION, NEW Y

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:DECUSATIS, CASIMER;GREGG, THOMAS A.;REEL/FRAME:019889/0814

Effective date: 20070920

STCF Information on status: patent grant

Free format text: PATENTED CASE

REMI Maintenance fee reminder mailed
FPAY Fee payment

Year of fee payment: 4

SULP Surcharge for late payment
MAFP Maintenance fee payment

Free format text: PAYMENT OF MAINTENANCE FEE, 8TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1552)

Year of fee payment: 8